Solar cells convert sunlight into electricity. These electronic devices have been traditionally fabricated using silicon (Si) as a light-absorbing, semiconducting material in a relatively expensive production process. To make solar cells more economically viable, solar cell device architectures have been developed that can inexpensively make use of thin-film, light-absorbing semiconductor materials such as copper-indium-gallium-sulfo-di-selenide, Cu(In,Ga)(S,Se)2, also termed CI(G)S(S). This class of solar cells typically have an absorber layer sandwiched between an electrode layer and a junction partner layer. The electrode layer is often Mo, while the junction partner is often CdS. A transparent conductive oxide (TCO) such as zinc oxide (ZnOx) is formed on the junction partner layer is typically used as a transparent electrode. CIS-based solar cells have been demonstrated to have power conversion efficiencies exceeding 19%.
A central challenge in efficiently constructing a CIGS-based solar cell is that the components of the CIGS layer must be within a narrow stoichiometric ratio in order for the resulting cell to be highly efficient. Achieving precise stoichiometric composition over relatively larger substrate areas is however difficult using traditional vacuum-based deposition processes. For example, it is difficult to deposit compounds and/or alloys containing more than one element by sputtering or evaporation. Both techniques rely on deposition approaches that are limited to line-of-sight and limited-area sources, tending to result in poor surface coverage. Line-of-sight trajectories and limited-area sources can result non-uniform three-dimensional distribution of nanoparticles in all three dimensions and/or poor film-thickness uniformity over large areas. These non-uniformities can occur over the nano-meso, and/or macroscopic scales. Such non-uniformity also alters the local stoichiometric ratios of the absorber layer, decreasing the potential power conversion efficiency of the complete device.
Alternative approaches to vacuum-based deposition techniques such as sputtering and evaporation have been developed. In particular, production of solar cells on flexible substrates using semiconductor printing technologies provides a highly cost-efficient alternative to conventional vacuum-deposited solar cells. For example, T. Arita and coworkers [20th IEEE PV Specialists Conference, 1988, page 1650] described a screen printing technique that involved mixing and milling pure Cu, In and Se powders in the compositional ratio of 1:1:2 and forming a screen printable paste, screen printing the paste on a substrate, and sintering this film to form the compound layer. They reported that although they had started with elemental Cu, In and Se powders, after the milling step the paste contained the CuInSe2 phase. However, solar cells fabricated on the sintered layers had very low efficiencies because the structural and electronic quality of these absorbers were poor.
Screen-printed CuInSe2 deposited in a thin-film was also reported by A. Vervaet et al. [9th European Communities PV Solar Energy Conference, 1989, page 480], where a CuInSe2 powder was used along with Se powder to prepare a screen printable paste. Layers formed by screen printing were sintered at high temperature. A difficulty in this approach was finding an appropriate fluxing agent for dense CuInSe2 film formation. Solar cells made in this manner also had poor conversion efficiencies.
U.S. Pat. No. 5,985,691 issued to B. M. Basol et al describes another particle-based method to form a Group IB-IIIA-VIA compound film. The described method includes the steps of preparing a source material, depositing the source material on a base to form a precursor, and heating the precursor to form a film. In that method the source material includes Group IB-IIIA containing particles having at least one Group IB-IIIA phase, with Group IB-IIIA constituents present at greater than about 50 molar percent of the Group IB elements and greater than about 50 molar percent of the Group IIIA elements in the source material. The powder is milled to reduce its particle size and then used in the preparation of an ink which is deposited on the substrate in the form of a precursor layer. The precursor layer is then exposed to an atmosphere containing Group VIA vapors at elevated temperatures to convert the film into the compound. The precursor films deposited using this technique were porous and they yielded porous CuInSe2 layers with small-grain regions as reported by G. Norsworthy et al. [Solar Energy Materials and Solar Cells, 2000, vol. 60, page 127]. Porous solar cell absorbers yield unstable devices because of the large internal surface area within the device, and small grains limit the conversion efficiency of solar cells. Another key limitation of this method was the inability to effectively incorporate gallium into the material. The properly-distributed presence of gallium in a CIS film serves to potentially broaden the bandgap of the semiconductor material, thereby increasing the open circuit voltage of the solar cell, and to promote the adhesion of the CIGS layer to a (Mo) electrode, providing a back surface electric field which can improve the collection of carriers. The absence of gallium decreases the potential power conversion efficiency of the solar cell. In practice, while gallium oxide particles can easily be produced, it is very difficult to reduce gallium oxide, even at relatively high temperatures, and in the absence of reduction, gallium oxide cannot be effectively used as a precursor material for gallium in the final film. Accordingly, in addition to poor stability, solar cells made using the approach of Basol et al. had sub-optimal power conversion efficiency.
Eberspacher and Pauls in U.S. Pat. No. 6,821,559 describe a process for making phase-stabilized precursors in the form of fine particles, such as sub-micron multinary metal particles, and multi-phase mixed-metal particles comprising at least one metal oxide. The preparation of particulate materials was described using a range of methods including laser pyrolysis, atmospheric arc evaporation, solution precipitation, chemical vapor reactions, aerosol pyrolysis, vapor condensation, and laser ablation. In particular, aerosol pyrolysis was used to synthesize mixed-metal particulates comprising metal oxides formed as substantially solid and spherical particulates. These particulate precursor materials were then deposited onto large-area substrates in thin layers using any of a variety of techniques including slurry spraying methods such as pneumatic spraying with a pressurized gas nozzle, hydraulic spraying with a pressurized slurry expelled through an orifice, and ultrasonic spraying with a rapidly vibrating atomization surface. A disadvantage of solar cell devices comprised of thin-film absorber layers formed in this manner was the poor reproducibility of the resulting device performance, and the porous form of the absorber layer, which tends to result in poor device stability.
Bulent Basol in U.S. Published Patent application number 20040219730 describes a process of forming a compound film including formulating a nano-powder material with a controlled overall composition and having particles of one solid solution. The nano-powder material is deposited on a substrate to form a layer on the substrate, and this layer is reacted in at least one suitable atmosphere to form the compound. According to one preferred embodiment of that process, the compound film has a Cu/(In+Ga) compositional range of 0.7-1.0 and a Ga/(In+Ga) compositional range of 0.05-0.3. Due to the improved process window made available by the phase space of a solid solution, the use of nanoparticles comprised of a solid solution may improve the repeatability and the overall yield of the thin-film deposition and solar cell production process. However, a means to incorporate additional Ga beyond that possible through a solid-solution (containing either Cu+Ga or In+Ga) restricts the potential performance of a device constructed by this method. In particular, since the presence of additional gallium in the light absorbing film serves both to broaden the bandgap of the semiconductor material and to increase the open circuit voltage of the solar cell, a lack of additional gallium in the light-absorbing thin film tends to decrease the potential power conversion efficiency of solar cells created in this manner. Using the solid-solution approach, Gallium can be incorporated into the metallic dispersion in non-oxide form—but only with up to approximately 18 relative atomic percent (Subramanian, P. R. and Laughlin, D. E., in Binary Alloy Phase Diagrams, 2nd Edition, edited by Massalski, T. B. 1990. ASM international, Materials Park, Ohio, pp 1410-1412; Hansen, M., Constitution of Binary Alloys. 1958. 2nd Edition, McGraw Hill, pp. 582-584.). However, efficient CIGS solar cells benefit from achieving a gallium ratio of up to 25 relative atomic percent. Furthermore, it would be simpler to directly work with elemental metallic nanoparticles rather than solid-solution metallic nanoparticles in that the elements can be optimized individually and they are more readily available in elemental form. However, no technique was known in the prior art to create gallium nanoparticle powders sufficient and adequate for semiconductor applications, in part because gallium is molten at room temperature and therefore does not lend itself to common techniques for creating nanoparticles in the form of powders that are then dispersed in solution (as commonly done with the other elements). As a result, it was not possible in the prior art to directly incorporate gallium (or incorporate gallium in a high percentage) into the metallic dispersion used to print the CIG precursor of a CIGS solar cell.
Thus, there is a need in the art for a method of forming a material comprised of gallium-containing CIGS precursor materials, where the precursor materials can be reproducibly, uniformly, and densely applied over large substrate areas to form a thin-film CIGS solar cell, and where the gallium is directly included in an elemental form.